Assessing indicators of agricultural drought impacts on spring wheat yield and quality on the Canadian prairies

Abstract

Drought is one of the costliest weather related natural hazards in the world. Traditionally, drought has been monitored using drought indices such as the “ percent of normal precipitation” to provide policy makers with information on drought extent and severity so that timely intervention measures can be taken. This study was conducted to identify drought indices that strongly correlated with spring wheat yield and quality. Detailed crop and weather data were collected from a series of wheat experiments across Saskatchewan and Manitoba, providing a diverse range of wheat growing environments. The data collected from 2003 to 2006 provided 15 site-years of data, which were used to derive drought indices in four categories: (i) water supply, (ii) water demand, (iii) water balance and (iv) water use. The drought indices were accumulated either over selected periods of the crop growing season including 1-month, 2-month, 3-month and 4-month (May–August) and/or specific wheat growth periods i.e., from planting to anthesis, anthesis to maturity and planting to maturity. Drought indices explained from 27% to 74% ( P < 0.05) of the variation in wheat grain yield and four different wheat bread-making quality parameters for two different wheat genotypes (AC Barrie and Superb). The drought index category most highly correlated to the greatest number of wheat characteristics was water demand, followed by water balance, water use and water supply. These results imply that water demand (evapotranspiration, ET) and water balance provide the most accurate assessment of drought impacts on spring wheat yield and quality on the Canadian prairies. Consequently, water demand and water balance indices would be more useful in assessing agricultural drought than the water supply indices frequently used for meteorological drought. The Hargreaves et al. method of estimating ET produced indices with correlation coefficients similar to those of the more complex modified Penman–Monteith method. Drought indices accumulated from planting to anthesis were slightly superior to those accumulated over the entire growing season, but highly superior to those accumulated from anthesis to maturity. Based on these results it would be possible to generate preliminary estimates of drought impacts on spring wheat yield and quality by mid-season with some confidence. This could be very useful in regards to providing policy makers and grain merchants with early warning information to plan and prepare for potential drought conditions.